ggplot2 Extensions

• Brief overview of ggplot2 + extensions
• Resources for extensions
• Some existing extensions
• Writing your own extension

Package for Grammer of Graphics

You provide the data, tell ggplot2 how to map variables to aesthetics, what graphical primitives to use, and it takes care of the details.

library(ggplot2)
library(dplyr)
library(gapminder)

pop_by_year <- gapminder %>%
  filter(country == "United States") %>%
  ggplot(aes(x = year, y = pop, group = country)) +
  geom_line()

pop_by_year

Completely customizable aesthetics

pop_by_year +
  labs(x = "Year", y = "Population", title = "US Population by Year") +
  scale_y_continuous(labels = function(x) paste0(round(x / 1e6), " mil")) +
  theme(axis.text.x = element_text(face = "bold"),
        panel.background = element_blank(),
        panel.grid = element_line(color = "blue"))

Easy visualization of data

gapminder %>%
  filter(continent == "Americas") %>%
  ggplot(aes(x = year, y = pop, group = country)) +
  geom_line() +
  facet_wrap(~country, nrow = 3) +
  scale_y_log10()

• https://ggplot2.tidyverse.org/
• http://r4ds.had.co.nz/data-visualisation.html
• http://www.cookbook-r.com/Graphs/
• https://www.rstudio.com/wp-content/uploads/2016/11/ggplot2-cheatsheet-2.1.pdf

• What if ggplot2 doesn't have the geom you want?
• extensions allow you to write your own geom
  • share with others in an R package
• http://www.ggplot2-exts.org/

life_exp_8707 %>%
  ggplot(aes(y = country, x = year_1987, xend = year_2007)) +
  ggalt::geom_dumbbell(colour_x = "red", colour_xend = "blue", size_x = 2, size_xend = 2)

• patchwork
• gganimate
• ggridges
• ggrepel
• ggExtra
• gghighlight
• ggthemes
• ggiraph

Combine plots together into one image

The goal of patchwork is to make it ridiculously simple to combine separate ggplots into the same graphic. As such it tries to solve the same problem as gridExtra::grid.arrange() and cowplot::plot_grid but using an API that incites exploration and iteration.

https://github.com/thomasp85/patchwork

gdp_plot + pop_plot

# gdp_plot + pop_plot - life_dist_plot + plot_layout(ncol = 1)
(gdp_plot | pop_plot) / life_dist_plot

gdp_plot + pop_plot - life_dist_plot + plot_layout(ncol = 1, heights = c(2, 1)) & theme_bw()

Turn plots into animated gifs

The core of the approach is to treat “frame” (as in, the time point within an animation) as another aesthetic, just like x, y, size, color, or so on. Thus, a variable in your data can be mapped to frame just as others are mapped to x or y.

https://github.com/dgrtwo/gganimate

• No longer a single ggplot image
  • doesn't play nice with markdown
• To make smooth transitions between your data use the tweenr package
• package is undergoing a total rewrite
  • aes(frame = ...) is going away at some point

Stacked density charts

Ridgeline plots are partially overlapping line plots that create the impression of a mountain range. They can be quite useful for visualizing changes in distributions over time or space.

https://github.com/clauswilke/ggridges

library(ggridges)
gapminder %>%
  ggplot(aes(x = gdpPercap, y = continent)) +
  ggridges::geom_density_ridges2() +
  scale_x_log10()

gapminder %>%
  ggplot(aes(x = gdpPercap, y = continent, fill = ..x..)) +
  ggridges::geom_density_ridges_gradient(scale = 3, rel_min_height = .01) +
  scale_fill_viridis_c() +
  scale_x_log10()

gapminder %>%
  ggplot(aes(x = gdpPercap, y = continent, fill = ..x..)) +
  ggridges::geom_density_ridges_gradient(scale = .9, rel_min_height = .01, position = position_raincloud(), jittered_points = T) +
  scale_fill_viridis_c() +
  scale_x_log10()

Prevents text labels from overlapping

ggrepel provides geoms for ggplot2 to repel overlapping text labels:

https://github.com/slowkow/ggrepel

library(ggrepel)

gapminder %>%
  filter(year == 2007, continent == "Americas") %>%
  ggplot(aes(x = gdpPercap, y = lifeExp)) +
  geom_point() +
  geom_text(aes(label = country))

library(ggrepel)

gapminder %>%
  filter(year == 2007, continent == "Americas") %>%
  ggplot(aes(x = gdpPercap, y = lifeExp)) +
  geom_point() +
  geom_text_repel(aes(label = country))

Add marginal density plots to scatter plots

https://github.com/daattali/ggExtra

p <- ggExtra::ggMarginal(life_vs_gdp_plot)

grid::grid.newpage()
grid::grid.draw(p)

p2 <- gapminder %>%
  ggplot(aes(x = gdpPercap, y = lifeExp, color = continent)) +
  geom_point() +
  scale_x_log10()

p2 <- ggExtra::ggMarginal(p2, groupFill = T)

grid::grid.newpage()
grid::grid.draw(p2)

Highlight relevant portions of your data

https://github.com/yutannihilation/gghighlight

• The package on cran is old

gapminder %>%
  ggplot(aes(x = year, y = gdpPercap, group = country)) +
  geom_line() +
  scale_y_log10()

library(gghighlight)

gapminder %>%
  ggplot(aes(x = year, y = gdpPercap, group = country, color = country)) +
  geom_line() +
  scale_y_log10() +
  gghighlight(country %in% c("United States", "Canada", "Mexico"), use_group_by = F)

gapminder %>%
  filter(continent == "Americas") %>%
  ggplot(aes(x = year, y = gdpPercap, group = country, color = country)) +
  geom_line() +
  scale_y_log10() +
  facet_wrap(~country, nrow = 3) +
  gghighlight(T, use_group_by = F, use_direct_label = F) +
  scale_color_discrete(guide = F)

Provide themes for ggplots

https://cran.r-project.org/web/packages/ggthemes/vignettes/ggthemes.html

http://davidgohel.github.io/ggiraph/articles/offcran/using_ggiraph.html

https://ggplot2.tidyverse.org/articles/extending-ggplot2.html

Will need a basic understanding of Object Oriented programming

gdp_plot

class(gdp_plot)

[1] "gg"     "ggplot"

?ggplot2::Geom

All geom_ functions (like geom_point) return a layer that contains a Geom object (like GeomPoint). The Geom* object is responsible for rendering the data in the plot.

• geom_point => GeomPoint object
• calling the function tells ggplot what data to use to create the object

To make your own extension you'll need to create your own Geom object (or Stat object if creating a stat layer)

• ggproto is the sytem of OO that ggplot uses.
• Different object types for different parts
  • Geom
  • Stat
  • Position
  • Scale
  • Coord
  • Facets
• Each type has its own methods you must customize (override)

• We need to tell ggplot what calculation to run and what geom to plot once we get the results

?Stat

StatMiddle <- ggproto("StatMiddle", Stat,
  compute_group = function(data, scales) {
    xmed <- median(data$x)
    ymed <- median(data$y)

    data.frame(x = xmed, y = ymed)
  },

  required_aes = c("x", "y")
)

Now we have a calculation that does something, we need to be able to call it like stat_smooth

stat_middle <- function(mapping = NULL, data = NULL, geom = "point",
                       position = "identity", na.rm = FALSE, show.legend = NA, 
                       inherit.aes = TRUE, ...) {
  ggplot2::layer(
    stat = StatMiddle, data = data, mapping = mapping, geom = geom, 
    position = position, show.legend = show.legend, inherit.aes = inherit.aes,
    params = list(na.rm = na.rm, ...)
  )
}

gapminder %>%
  ggplot(aes(x = lifeExp, y = gdpPercap)) + 
  geom_point() +
  scale_y_log10() +
  stat_middle(color = "red", size = 2)

gapminder %>%
  ggplot(aes(x = lifeExp, y = gdpPercap)) + 
  geom_point() +
  scale_y_log10() +
  stat_middle(aes(color = continent), size = 2)

https://github.com/AtlantaRUsers/Meetups